A biotech company, GRAIL, Inc., has developed a blood test capable of detecting 50+ types of cancer, even in people without signs of symptoms. In a large trial of over 23,000 people in the U.S. and Canada, all aged 50 and older and none showing symptoms, the test identified cancer signals in 216 people. Of those, 133 were confirmed to have cancer, a positive predictive value of about 61.6% in that cohort.
More than half of the cancers the test detected were in early stages (Stage 1 or 2), including many cancers for which no standard screening currently exists. The false positive rate, which tells us how often the test flags cancer by mistake, was very low in that early data set, at about 0.4%. Using AI algorithms, the test also correctly determined the tissue or organ of origin in about 92% of confirmed cases in the study.
How the test works
Your blood carries tiny bits of DNA from cells that die every day. If a person has cancer, some of those bits come from cancer cells. This is called circulating tumor DNA (ctDNA). The test looks for three clues:
- DNA patterns (methylation): Cancer cells change the way certain genes are switched on or off by adding tiny chemical tags called methyl groups to the DNA. These methylation patterns act like fingerprints. Each cancer type leaves its own distinct mark that the test can recognize.
- DNA shapes (fragmentation): When cells die, their DNA breaks into pieces. In cancer, these fragments are often shorter or cut in unusual ways. The test studies these DNA “shapes” to spot irregular patterns linked to cancer.
- Where it came from (tissue of origin): A computer algorithm (AI) compares the DNA pattern to thousands of known profiles from different organs. This helps the system predict which body part, like the lung, liver or breast, the DNA most likely came from.
In practice, a blood sample is collected, DNA is extracted from it, and that DNA is analyzed for methylation patterns and fragmentation. If no cancer signal is detected, the person carries on with regular monitoring. If a signal is detected, the test predicts the tissue or organ of origin, the finding is confirmed clinically with imaging or a biopsy, and the result is an early diagnosis and treatment planning.
Why it matters
Detecting cancers at Stage 1 or 2 means much higher survival and more treatment options than late-stage disease. If further approved, this test could significantly shift the cancer survival rate. Many cancer types, such as cancers of the pancreas, liver and ovaries, currently lack routine screening, and a broader test like this fills a major gap.
If widely adopted, this blood test may become the standard way to detect cancer in the mass population. Early-stage intervention could also reduce downstream treatment costs and shift investment toward initial diagnostics and early detection. And the test could give underserved populations and areas convenient, affordable access.
The data is promising, but full regulatory approval (for example by the U.S. Food & Drug Administration) and large-scale real-world deployment tests remain ahead. Watch for more data in the future, including results from larger and more diverse population pools and more clinical applications.